TestSetup

Exploratory analysis of series with a potential to nowcast GDP

Goals

• Plot series used in the Index of the Canadian Economic Dashboard and COVID-19
• Identify data tables and vectors for the dashboard

Comments

• Time series from Canadian Economic Dashboard and COVID-19 at https://www150.statcan.gc.ca/n1/pub/71-607-x/71-607-x2020009-eng.htm
• There are 16 indicators

suppressMessages(library(cansim))
suppressMessages(library(ggplot2))
suppressMessages(library(dplyr))
suppressMessages(library(plotly))
source("../../src/utils.R")

Extract the Data

vector_names <- c(
  "v65201210", "v41690973", "v2062809", "v1001827265",
  "v1001826653", "v52367097", "v4391505", "v800450",
  "v32858858", "v32858865", "v32858872", "v74869", "v129449",
  "v129460", "v129472", "v129482"
)
length(vector_names)

[1] 16

vector_descr <- c(
  "RealGDP", "CPI", "Employment [pers]", "Int. merchandise trade Exp. [$]",
  "Int. merchandise trade Imp. [$]", "Retail Sales [$]", "hours worked",
  "Manufact. sales [$]", "Aircraft domestic [#]", "Aircraft transborder [#]",
  "Aircraft int other [#]", "Railway carloads [tons]", "Travelers US [pers]",
  "Travelers other country [pers]", "CA resident US [pers]",
  "CAresident other country [pers]"
)
names(vector_descr) <- vector_names

I expect this date to be before the start of all time series/vectors.

start_date <- "1980-01-01"

Retrieve all vectors. They are concatenated along axis 0.

vectors <- get_cansim_vector(vector_names, start_date)
dim(vectors)

[1] 6315    9

Preprocessing

colnames(vectors) <- sapply(colnames(vectors), tolower)
colnames(vectors)

[1] "decimals"      "value"         "ref_date"      "releasetime"   "symbol"        "frequencycode"
[7] "scalar_id"     "coordinate"    "vector"       

vectors$ref_date <- as.Date(vectors$ref_date)

head(vectors)

vector_names[1]

[1] "v65201210"

length(vector_names)

[1] 16

indicators1 <- vector(length = 16)
for (i in 1:length(vector_names)) {
  indicator <- filter(vectors, vector == vector_names[i])
  ref_value <- indicator$value[indicator$ref_date == as.Date("2019-12-01")]
  indicator$value_index <- 100. * indicator$value / ref_value
  if (i == 1) {
    indicators <- indicator
    }
  else {
    indicators <- rbind(indicators, indicator)
    }
}

head(indicators)

indicators$Descr <- sapply(indicators$vector, function(x) {
  vector_descr[x][[1]]
})

Canadian Economic Dashboard Visualization

p <- ggplot(indicators, aes(x = ref_date, y = value_index)) +
  geom_line((aes(group = Descr, color = Descr))) +
  scale_x_date(date_breaks = "3 month", date_labels = "%b %Y",
               limits = as.Date(c("2019-01-01", "2020-09-01"))) +
  scale_y_continuous(breaks = seq(0, 180, 20), limits = c(0, 180)) +
  labs(title = "Index, December 2019 = 100", x = "", y = "")
ggplotly(p)

Convert Dashboard series into a tsibble for forecasting

• Start Date = January, 2000

head(dashboard)

Plot series

as_tsibble(filter(indicators, ref_date >= "2000-01-01")%>%select(value,ref_date,Descr)%>%mutate(Month=yearmonth(ref_date),Series=Descr)%>%select(Month,value,Series),index=Month,key=Series)%>%mutate(Date=as.Date(Month))->SeriesData1
p <- ggplot(SeriesData1, aes(x = Date, y = value)) +
     geom_line((aes(group = Series, color = Series))) +
     scale_x_date(date_breaks = "2 year", date_labels = "%Y",
                  limits = as.Date(c("2000-01-01", "2020-09-01"))) +
     labs(title = "Series for Forecasting", x = "", y = "")
ggplotly(p)